Does high dietary fructose cause hypertension?

Average fructose intake in the U.S. is staggering: Americans currently consume approx. 150 pounds of sugar a year, which is 20 times the amount their European ancestors would have ingested from fruit (Data from USDA). This intake is mainly due to added sugar, high-fructose corn syrup (HFCS) in particular, in sugary soft drinks and bakery products. HFCS was introduced in the 1960s and, since then, annual per capita consumption has increased from 0 to 29 kilograms in 2001, while sucrose intake has decreased from 44 to 30 kg.


This increase in fructose consumption correlates with a rise in the prevalence of hypertension and evidence from animal models suggests fructose can raise BP. In a cross-sectional study of 4,528 participants from NHANES without a history of hypertension, dietary fructose ≥ 74 g/d (corresponding to 2.5 sugary soft drinks) was independently associated with elevated SBP after adjusting for demographics, comorbidities, physical activity, total kilocalorie intake, and dietary confounders such as total carbohydrate, alcohol, salt, and potassium intake. Furthermore, in a prospective analysis of 810 adults by Chen et al. cutting out one sugar-sweetened beverage per day was associated with a 1.8/1.1-mmHg reduction in BP over 18 months after controlling for potential confounders.


How fructose might raise BP is not known, but an interesting hypothesis implicates uric acid metabolism. Unlike other sugars, fructose causes an increase in the production of uric acid in animal models. Hyperuricemia increases juxtaglomerular renin production and decreases macula densa neuronal NO synthase expression, causing renal vasoconstriction, sodium retention and increased BP. Persistent renal vasoconstriction and preglomerular microvascular disease leads to the development of salt-sensitive HTN, even after hyperuricemia is corrected. In addition, uric acid may induce vascular smooth muscle cell proliferation and preglomerular arteriopathy following its uptake via probenecid-sensitive urate-transport channels (URAT1) in vascular smooth muscle cells.


Although the investigators in the NHANES study above adjusted for diet, residual confounding may also explain the association between HFCS and HTN i.e. soft drinks go hand in hand with a burger and fries, leading to salt-loading. So, along with a low-salt diet or DASH, is it time to start recommending low-fructose diets to delay pre-hypertension, incident hypertension and metabolic syndrome?


Boonsong Kiangkitiwan M.D.

Warfarin-induced AKI

We have recently posted about problems using warfarin in renal patients. On a similar note, I was recently involved in the care of a 58-year-old African-American patient who presented with oliguric AKI and a supratherapeutic INR level of 14.9 while on warfarin therapy for pulmonary embolism. Although AKI was probably secondary to ischemic ATN from massive bleeding (requiring 11 units of PRBCs), this patient made me wonder whether warfarin could itself cause AKI in the absence of significant hemodynamic changes.

It has been suggested that warfarin may induce AKI by causing glomerular hemorrhage and renal tubular obstruction by RBC casts. The first report by Abt, et al observed glomerular hematuria in a patient with excessive warfarin anticoagulation with underlying structural abnormality of glomerular basement membrane, suspected warfarin-induced glomerular hematuria. This finding was also noted by Kabir et al and Brodsky et al. subsequently reviewed renal biopsies from 9 patients with elevated serum creatinine level (mean, 4.3±0.8 mg/dL) and abnormal INR (mean, 4.4±0.7 IU) after excluding acute/active glomerulonephritis. INR was normalized before the kidney biopsy. The biopsy specimens revealed RBCs in both Bowman spaces and tubules. The occlusive RBC casts were mostly localized to distal nephron segments. The presence of RBCs was unlikely related to the biopsy procedure as the distribution of RBCs was within the nephron, non-uniform, and not in the interstitium or around the edges of the biopsy specimens. Moreover, the presence of occlusive RBC casts compressing tubular epithelium suggested glomerular hematuria. All of these patients had underlying kidney disease. Recovery was incomplete in six of nine patients. Based on this review, it seems that warfarin is unlikely to cause AKI in patients with normal renal parenchyma. However, elderly patients with underlying kidney diseases may be at risk for developing warfarin-associated AKI by this mechanism.

In addition to this potential renal complication of warfarin therapy, there were 2 case reports of warfarin-induced AIN by Volpi et al. and Kapoor et al. The latter case reports a patient on warfarin for 2 months for DVT who presented with AKI, supratherapeutic INR and skin rash. Renal biopsy revealed AIN and punch skin biopsy showed LCV with high eosinophil count, highly suggestive of drug-induced reaction after a negative comprehensive work-up and the absence of other recent medication changes. Warfarin was withdrawn and the patient’s renal function began to resolve upon discharge. Spontaneous cholesterol embolization is also widely reported in patients with diffuse atherosclerosis on anticoagulation.

In summary, it appears that in patients with pre-exiting CKD, warfarin may precipitate AKI via either glomerular hemorrhage and intra-tubular obstruction, AIN, spontaneous cholesterol embolization, in addition to hemodynamic change secondary to massive blood loss. This should be borne in mind in cases of AKI in patients taking warfarin, when diagnosis is unclear.

Boonsong Kiangkitiwan, M.D.